Chromatin Models. Interactions between DNA and Polypeptides Containing L-Lysine and L-Valine: Circular Dichroism and Thermal Denaturation Studies' Richard Mandel* and Gerald D. Fasman*
ABSTRACT: The interaction of calf thymus D N A with statispolymers of L-LYSand L-Val to form Q-type complexes with tical copolymers of L-lysine and L-valine [poly(L-LYsIOO~-L- DNA, they bind as well to D N A as does poly( L - L ~ sand ) give Valf)] and block copolymers [poly(~-Lys)'~~~-poly(~-Val)~] rise to a thermal denaturation pattern showing bound peaks were investigated as a function of ionic strength using circular between 90 and 100 "C, seen clearly with C D denaturation. dichroism (CD) spectroscopy. It was found that valine supThe thermal denaturation of mixed and annealed complexes presses the ability of the copolymer-DNA complexes to yield of poly(~-Lys)-DNA shows similar patterns in hqperchroma *-type C D spectra as found for poly(~-Lys)-DNA[Jordan, icity changes as a function of temperature but very different C . F., Lerman, L. S., and Venable, J . N. (1972), Nature C D melts. From the C D melt of annealed poly(~-Lys)-DNA. (London),New Biol. 236, 671 and lowers the ionic strength at it appears that aggregation and long-range order of the comwhich C D distortion occurs. Thermal denaturation, simultaM EDTA) as well plex are significant in low salt (2.5 X neously monitoring 280-nm ellipticity, [8]280, and hyperas in 1 .O M NaC1. These studies further illustrate the imporchromicity, h280, was carried out on annealed complexes of tance of the nature of nonionic interactions (hydrophobic) poly(~-Lys)-DNA, p o l y ( ~ - L y 5-~-Va115.5)-DNA, s~~ between polypeptides and D N A in determining the behavior poly(~-Lys)~~.~-poly(~-Val)~~ 8-DNA and directly mixed of their complexes, such as causing condensation into higher M EDTA, pH complexes of poly(^-Lys)-DNA, in 2.5 X order asymmetric structures. In light of these observations, the 7.0 solution. The C D denaturation of uncomplexed D N A a t possible significance to the C D melting of chromatin and the several ionic strengths was also determined to examine prevalidity of identification of C-form D N A by C D spectroscopy melting. Despite the inability of both statistical and block coare discussed.
c o m p l e x e s between p ~ l y ( ~ - L yand ~ ) 'DNA (L-DNA) have been widely used as a model for the interaction of histones with D N A (Olins et al., 1967; Shih and Bonner, 1970) and as a probe of the structure of chromatin (Itzhaki, 1971; Li, 1972; Clark and Felsenfeld, 1971; Umiel and Plaut, 1973). More recently, the interaction between D N A and random copolymers containing other amino acids in addition to Lys, Le., Leu (Ong et al., 1976; Ong and Fasman, 1976), Ala (Sponar et al., 1973), as well as sequential copolymers containing L-LYS, L-Ala, L-Pro, L-Gly (Sponar et al., 1973, 1974), L - L ~ s~, - G l y (Brown et al., 1974), and L-LYSand L-Tyr (Santella and Li, 1974) have been studied as a means of understanding the role of peptide sequence and secondary structure on the strength. specificity, and mode of binding with DNA. It has been observed that p o l y ( ~ - L y ~forms ) qualitatively different complexes with DNA, depending on the salt con-
centration a t which binding occurs. When the complex is prepared by direct mixing at salt concentrations on the order of 1 .O X M, it is found that: the circular dichroism (CD) spectrum of the complex depends linearly on the ratio of Lys to DNA; the C D spectrum of Lys bound regions in the complex is similar to that of DNA in 6 M LiCl (Chang et at.. 1973). On the other hand when p o l y ( ~ - L y is ~ )bound at YaCl conccntrations of around 1 .O M, the resulting complex shows cooperative binding characterized by a very intense C D spectrum with a peak ellipticity of - 100 000 a t 270 nm (Shapiro et 2 1 . ~ 1969), C D changes which are nonlinear with respect to I . ) \ input, and significant light scattering a t wavelengths greater than 320 nm (Carroll, 1972). Similar optical properties havc also been seen in complexes between H 1 histone and D h / Z dialyzed down to 0.15 M NaCl (Fasman et al.. 1970: Sponar and Fric, 1972), films of L i - D h A and Li-d(A-T) a t 92% o r lower relative humidity (Brunner and Maestre, 1974) and films of Na-DNA at 75% humidity (Tunis-Schneider and Maestre. This is publication No. 1089 from the Graduate Department of Biochemistry, Brandeis University. Waltham, Massachusetts 02 154. Receiced 1970) and D N A in sohtions of NaCl and polyethylene oxide October 24. 1975. This research was generously supported in part by grants (the \k form) (Jordan et al.. 1972). from the U S . Public Health ( G M 17533), National Science Foundation Following verification that the three predominant canonical (GB29204X), and the American Cancer Society (P-577). forms of DNA, A, B, and C , as determined by x-ray crystal1 Postdoctoral trainee of the National Institutes of Health. General Medical Sciences Training Grant No. G"v1212. lography, have three distinct C D spectra (Tunis-Schneidcr and * John Simon Guggenheim Fellow 1 9 7 4 1 9 7 5 . Maestre, 1970), C D spectroscopy has been utilized to stud! I Abbreviations used: CD, circular dichroism; L-DNA, poly-i.-lysine conformational changes in DNA: in salt solutions (Ivanoi, et DNA; block LV-DNA, p o l y ( ~ - l y s i n e *-poly(L-valine)'* )~~ X-DNA complex; statistical L V - D h A. p ~ l y ( L - l y s i n e ~ ~ ~ ~ - ~ -5)-DNA v a l i n e ~ ~ al., 1973; Zimmer and Luck, 1973), as a function of tempercomplex; T,f. maximum of derivative of hyperchromicity with respect ature prior to melting (Gennis and Cantor, 1972; Studdert et to temperature of unbound segments of DNA; T , I , maximum of derivaal., 1972), in complexes of polypeptides with DNA (Chang ci tive of hyperchromicity with respect to temperature of bound segments al., 1973), in complexes of histones with D N A (Fasman et al.. of DNA; T,z, maximum of derivative of hyperchromicity with respect 1970; Shih and Fasman, 197 1 ) and in chromatin (Hanlon et to temperature of bound segments of D N A ; EDTA, ethylenediamineteal., 1972; Shih and Fasman, 1970). This paper examines the traacetic acid; NalEDTA, disodium ethylenediaminetetraacetate; Cbz. benzyloxjcarbonyl; dp, degree of potymerkation. optical properties of gradient dialyzed complexes bet%eeri +
~
3122
BIOCHEMISTRY. VOL.
15,
NO.
14, 1976
LYS-VAL
TABLE I:
POLYPEPTIDE: DNA COMPLEXES
Copolymers of L-LYSand L-Val. Polymerization Conditions, Amino Acid Ratios, and Molecular Weights. Polymer
Lys:Val’ Ratio
G F-21-353-25 G F-21-329-23
loo.o:o.o loo.o:o.o
GF-23-48-17 G F-23-44-24 GF-23-142-25 GF-23-448-1 5e
94.4:5.6 86.k13.2 85.5:15.5 87.2:12.8
AIIb
[qlg
Mol Wtd
dP
200 200 25 25 25 20/
0.81 1.48 2.98 1.16
53 000 106 000 216 000 80 000 141 OOOc 124 OOOc
322 645 1340 512 950 756
C
a Determined by amino acid analysis. Anhydride to initiator ratio. Not soluble enough for viscosity measurement. Molecular weight estimated from urea gel electrophoresis (Walter and Fasman, to be published). Estimated from calibration curve for poly-L-lysine(Applequist, 1959). e Block copolymer, poly(~-Lys)~~.~-poly(~-Val)~~.~. f n-Hexylamine initiator for the Nf-Cbz-L-lysine-N-carboxyanhydride. g 1 M NaCl pH 4.0.
random and block copolymers containing Lys and Val [poly(~-Lys)~~~~-poly(~-Val)~] and calf thymus D N A as a function of ionic strength and compares these results with those of gradient dialyzed poly(~-Lys)-DNA(L-DNA) complexes a t the same lysyl to phosphate ratios ( r ) . Thermal denaturation is a useful tool to determine the extent and strength of binding of D N A to polypeptides (Li, 1973) and histones (Ansevin and Brown, 1971) and as a probe of nucleohistone structure (Subirana, 1973). This method has been extended in a previous paper (Mandel and Fasman, 1974) by simultaneous measurement, a t a single wavelength, of both absorption and C D changes prior to and during melting. At temperatures below melting, the ellipticity a t 280 nm of D N A is very sensitive to temperature (known as premelting) (Gennis and Cantor, 1972), and is thought to reflect conformational changes in the D N A double helix structure. In addition, ellipticity changes prior to and during thermal denaturation have been followed in complexes of H 4 and D N A (Adler et al., 1975), and in chromatin (Henson and Walker, 1970; Wilhelm et al., 1974). T o better understand the significance of premelting, the thermal denaturation of DNA, monitored by simultaneous absorption and C D measurement, was carried out over a wide range of salt concentration. In addition this method has been applied to both directly mixed and annealed complexes of p o l y ( ~ - L y ~with ) D N A as a probe of structural differences between these two distinct complexes and as models for the understanding of the significance of the premelting phenomena observed in other polypeptide and protein complexes with DNA, as well as in chromatin. Furthermore, the thermal denaturation of annealed complexes between the statistical copolymer p o l y ( ~ - L y s ~ ~ ~ ~ -and ~-V thea block l ~ ~ ~copolymer ~) poly(~-Lys)~~.~-poly(~-Val)~~.~ with D N A has been studied to examine the role that hydrophobic residues play in interactions between proteins and D N A and their effect on the structure and physical properties of the complex. Experimental Section
Calf thymus DNA was prepared as previously described (Fasman et al., 1970) and had a molecular weight of approximately 10- 12 X 10-6 as determined by ultracentrifuge sedimentation. Poly-L-lysine was prepared as previously described (Fasman et al., 1961) with intrinsic viscosity and estimated molecular weight shown in Table l , Random copolymers of L-lysine and L-valine [poly(Ll y ~ i n e ~ ~ ~ ~ - ~ - v awere l i n esynthesized ?] as previously described (Mandel and Fasman, 1975).
Block copolymer, (L- Lys 87,2),, (L- Leu 2,8), Synthesis. NWbz-L-lysine-N-carboxyanhydride (0.603 g) (Fasman et al., 1961) was dissolved in dry dioxane (61 ml) and 0.328 ml of 0.300 N n-hexylamine, in benzene, was added to initiate polymerization (anhydride/initiator = 20). The solution became viscous after standing 5 days. L-Valine-N-carboxyanhydride (0.0497 g) (Kubota and Fasman, 1975), dissolved in freshly distilled benzene (5 ml), was added to the above viscous solution which was stirred for 5 min and allowed to stand for 4 days. The viscous solution was diluted with CHC13 (60 ml), stirred well, and treated with HCl gas and HBr gas (30 min each) in the manner described by Fasman et al. (1961). After stirring for 2 h, the white precipitate which had formed was allowed to settle and the supernatant was removed by suction. The precipitate was dissolved in H2O (300 ml), the solution adjusted to p H 7 with 1 N HC1 and extracted twice with ether, the p H adjusted to 5.0, and the solution dialyzed in a fiber dialyzer (Bio-Rad, mini plant No. c/HFD-15, molecular weight cutoff 5000) for 4 h vs. H 2 0 , 4 h vs. 0.01 N HCl, and finally lyophilized (weight, 0.15 g). The amino acid compositions as well as intrinsic viscosities and estimated molecular weights are shown in Table I . Annealed complexes were prepared by addition of a 2 M NaCl solution of polypeptide to a 2 M NaCl solution of D N A to obtain the correct lysyl to phosphate ratio, with subsequent dialysis at 4 OC from 1.5 M NaCl to the required salt concentration. Both linear gradient and exponential gradient dialysis were used (Carroll, 197 1). Exponential gradient dialysis was performed at a rate such that the complexes went from 1.5 to 0.5 M salt over the course of at least 5 h. Dialysis bags were prepared by boiling in 5% NaHC03-2 X I O p 3 M NazEDTA, washing in double distilled water, and boiling with double-distilled water prior to use. The concentration of D N A stock solutions was determined by absorption measurement with 6260 6800. Concentration of p o l y ( ~ - L y ~stock ) solutions was determined by biuret determination (Zamenhof, 1957) and the L-LYS,~ - V acopolymer l concentrations by micro-Kjeldahl analysis (Lang, 1958). Due to volume changes during dialysis it was necessary to determine the D N A concentration in the complex. This was determined from OD,,, of 0.5 N HC104 hydrolysates (30 min in boiling water bath) (Adler et al., 1971) or by 1 : l dilution of the complex with 4.0 M NaCI. Both methods yield consistent results. Ultraviolet absorption spectra were measured in a Cary 14 recording spectrophotometer a t 23 O C . Circular dichroism spectra were measured in a Cary 60 recording spectropolarimeter with a Model 6001 C D attach-
I
BIOCHEMISTRY, VOL.
15,
NO.
14, 1 9 7 6
3123
IO0
A
85-
--
&--O-d\
\
/&--
P '
n I
0
* 2 70-
-
i
0
N
0
0
9
Y
,
D N A in NaCl
40
20
0
40
80
60
100
Temperature
I50
c
DNA i n N a C l
-loo
I t
-150
1
- 300 0
I
I
20
1
I
I
I
40 60 Temperature
1
80
I.o
0 5
0
0
I
I
N a C l Concentration 100
t I G L R t I Denaturationof DNA in 2 5 X IOT4 M EDTA, pH 7 0 ( 0 - 0 ) . 0 012 M N a C l + 2 5 X IO-4 M E D T 4 , pH 7 0 ( A - - A ) , 1 0 M YaCl 2 5 x 10-4 M EDTA, p~ 7 o ( 0 . .. . .. 0 ) ( A ) Ellipti~itj,[O]ixo. \ \ temperature (B) Derivative of ellipticity. [8]28o/dT. with respect to temperature
+
2: Ellipticity. [e]280.vs. N a C l concentration of poljpeptide DNA complexes. at r = 0.5. Poly(L-lysine)--DNA ( 0 - 0 ) ;poIy(i.-l\sine944 - ~ - v a ~ i n~e ') - D N A( A - - A ) ; poij(~-lysine*6g-L-valinei2 2 ) - D&;\ ( 0 . . . . 0):p o l y ( t . - l y ~ i n e )'-poly(t.-valine)'*.~ ~~ Dh'A (e-e.). FIGLRE
___-.._ .--
~
_-
..
of D h i l vs. Salt Concentration Measured by Ellipticitj [O]ZXOa n d Hyperchromicity a t 280 nni.
T IHI-L- 1 1 : Thermal Denaturation
ment as previously described (Adler et al.. 1971). Ellipticity. [ 01, is expressed in units of deg mol- I cm-' , Thermal denaturation measurements were carried out as previously described (Mandel and Fasman, 1974) on solutions with DNA concentrations of 1 .O-2.3 X IO-4 M phosphate. To facilitate comparison of different complexes which scatter light and therefore have significantly different extinction coefficients, the absorption denaturation curves are normalized and plotted as hyperchromicity derivatives vs. temperature dh
(E)i =
:i;:::y;)
where ( d h / d T ) i is the derivative of hyperchromicity with respect to temperature a t temperature corresponding to point i, i f 1 refers to data points adjacent to i, and to is the lowest measured extinction. C D denaturation curves are drawn as ellipticity, [ 0 ] 2 ~ 0 .The derivative of ellipticity with respect to temperature d [e]280/d T vs. temperature is useful for identification of thermal transitions and measurement of the premelting slope. The premelting slope is defined as the average value of d[0]280/dT between approximately 5 and 35 OC. Direct mixed complexes were prepared by slowly adding an equal volume of polypeptide at the correct concentration in 2.5 X M EDTA (pH 7.0) in a solution of DNA in 2 . 5 X M EDTA and dialyzing the resulting complex in a large volume of 2.5 X M EDTA (pH 7.0).
3124
BIOCHEMISTRY, ~
O
L1
5 , ho
14. 1916
as
haCi
T,,(absorption) f 1.5 "C
0.0007'~ 0.0120 1 .00
T,,,(eIlipticit>) f 1.5 oc
47.3
47.8
65.1
65.7
91.6
91.9
A\ Preinelr Slopeh
[0]3k0
'4.0 It x 19.0 f 8 25.4 f 8
9300 7900
7YOO
\ a + concentration corresponding to 2.5 X IO-' ILI EDTA. p t l /' The average value of premelt was determined b) aleraging thc Ic1Iues of (d[8]280/dT)1 for all measured points bctaeen approximatel> 5 and 35 O C
7.0.
Results
Prrnielting oj'f1:VA. In order to examine the salt dependerice of premelting, thermal denaturation of D N A monitored at 2x0 nm was carried out in 2.5 X M EDTA and 0.01 2 M and 1 .O M NaCl and is shown in Figure 1. The calculated average slope of the premelt, the highest attained ellipticity, and the T,,, from both absorption (not shown) and C D are summarized in Table 11. It can be clearly seen from the parallel lines at 1 o ~ temperature in Figure 1A that the slope of the premelt is independent of salt over a wide range of salt concentration. This is in general agreement with previous results obtained at 0 . 5 and 0.05 M NaCl (Gennis and Cantor, 1972). I t should also be noted that the maximum attained ellipticit>,as well as the
L.YS-VAL
POLYPEPTIDE: DNA COMPLEXES
Critical NaCl Concentration for Maximal Circular Dichroism Changes in Copolymer-DNA Complex.
TABLE I I I :
~~
____
LYS
Copolymer
r
Critical [NaCI]
0
[el280
~
0.5
0.5 Poly(~-Lys944, L-Vals 6 ) 0.5 Poly(~-Lys*~.8, ~ - V a l ’ 2) 3 0.5 P ~ l y ( L - L y s ) ~*’- p o l y ( ~ - V a I ) ~ ~ ,~
1.02 0.93 0.83
